OBJECTIVE To compare, in a clinical study of a special design, the pharmacokinetic profile of mirtazapine in 20 young healthy male volunteers on two treatment regimens with homothetic oral tablets at steady state: NOCTE (1 x 30 mg at 21.00 h) and BID (15 mg at 21.00 h and 15 mg at 09.00 h). METHODS Pharmacokinetic parameters were calculated from mirtazapine plasma levels assayed by gas chromatography with nitrogen-sensitive detection. A special analysis of variance allowed interesting interactions to be distinguished. RESULTS The steady state was reached after 4 and 6 days for NOCTE and BID respectively; the difference was presumably due to intersubject variability. In accordance with pharmacokinetic theory, the peak-to-trough ratio at steady state was significantly lower (twofold) for BID than for NOCTE. Within BID, a small difference (approx. 10%) was found in the extent of absorption between evening and morning administration. Although statistically significant, this difference meets strict bioequivalence requirements. The regimens NOCTE and BID were found to be bioequivalent for the steady-state area-under-the-curve-curve and the peak time. Bioequivalence testing for the peak level was not meaningful due to the difference in dosing regimens. The observed elimination half-lives of 19.7 +/- 3.0 h and 20.8 +/- 2.7 h (n = 20) for NOCTE and BID, respectively are in agreement with previous results. CONCLUSIONS Differences (if any) were found to meet strict bioequivalence requirements and were so small that they are of no clinical consequences.